Responses of legumes and grasses to non-, moderate,and dense shade in Missouri, USA. I. Forage yieldand its species-level plasticity

Kejia Pang . J. W. Van Sambeek . Nadia E. Navarrete-Tindall .

Chung-Ho Lin . Shibu Jose . H. E. Garrett

Received: 17 June 2016 / Accepted: 11 January 2017

� Springer Science+Business Media Dordrecht 2017

Abstract Annual screenings of forage grasses and

legumes for shade tolerance were conducted from

1996 to 2001 in the outdoor Shade Tolerance Screen-

ing Laboratory at the Horticulture and Agroforestry

Research Center, University of Missouri. Forty-three

forages were grown under non-shade (100% of full

sunlight), moderate shade (45%), and dense shade

(20%) without competition for water and nutrients.

Annual forage yield (g pot-1) was equal to or higher

under moderate shade for all 43 forages and under

dense shade for 31 forages than the non-shade control.

Relative distance plasticity index (RDPI), a measure

of a species’ adaptability to different environments,

ranged from 0.104 to 0.567. Cool season grasses had

the lowest RDPI (0.183), followed by warm season

grasses (0.252), warm season legumes (0.274), and

cool season legumes (0.314), indicating grasses tend to

be more shade tolerant than legumes in terms of forage

yield. Overall, most grass and legume forages have the

potential to produce equivalent or higher yields in

agroforestry practices featuring light to moderate

shade than forages in open pastures when competition

from tree roots is minimized.

Keywords Shade tolerance � Relative distance

plasticity index � Annual biomass � Warm-season

forages � C4 grasses

Introduction

An agroforestry system, in which trees and forages or

crops are intentionally integrated, has both economic

and ecological benefits including diversifying sources

of income for landowners, mitigating soil erosion,

improving water quality, increasing biodiversity, and

moderating microclimate extremes (Gold and Garrett

2009). The microclimate factors affected include

humidity, soil and air temperature, soil moisture, and

especially, light intensity and quality (Martsolf 1966;

Callaway 2007). Consequently, the yield of forages

grown in agroforestry systems are likely to differ from

forages grown in open fields or pastures (Pearson

1983; Watson et al. 1984). Our knowledge about the

Electronic supplementary material The online version ofthis article (doi:10.1007/s10457-017-0067-8) contains supple-mentary material, which is available to authorized users.

K. Pang (&) � C.-H. Lin � S. Jose � H. E. Garrett

Center for Agroforestry, University of Missouri, 203

Natural Resource Bldg., Columbia, MO 65211-7270,

USA

e-mail: pangkejia@gmail.com

J. W. Van Sambeek

Northern Research Station, USDA Forest Service, 202

Natural Resource Bldg., Columbia, MO 65211-7270,

USA

N. E. Navarrete-Tindall

Cooperative Extension, Lincoln University, 306 Allen

Hall Bldg., 900 Chestnut Street, Jefferson City,

MO 65101, USA

123

Agroforest Syst

DOI 10.1007/s10457-017-0067-8

shade tolerance of forage species is usually qualitative,

i.e., whether they occur in shaded environment, while

quantitatively how adaptive these forages are to

reduced light remains little explored.

Investigations of shade tolerance for forages have

been conducted under a wide range of environmental

conditions, from green shade created by tree canopies

to neutral shade produced by shade cloth or slats under

which the distribution of light wavelengths remain

unchanged. Examples include studies by Peri et al.

(2002) on the photosynthetic response of cocksfoot

(Dactylis glomerata) and white clover (Trifolium

repens) in a silvopastoral setting, by Beard (1965) on

qualitative differences in survival rate and visual

quality of eight turf grass species under tree canopies,

and by Varella et al. (2010) on alfalfa (Medicago

sativa) responses to reduced light under both shade

cloth/slats and tree canopies in agroforestry systems.

Both light intensity and quality (i.e., red: far-red ratio)

under tree canopies can vary depending on species,

canopy density, and duration of sun flecks (Martsolf

1966; Varella et al. 2010), but light intensity under

neutral shade is relatively easy to quantify and light

quality remains unchanged. Researchers have evalu-

ated the shade tolerance of forages using shade cloth

suspended over field-grown plants (Watson et al.

1984; Devkota et al. 1997; Jiang et al. 2004; Ehret

et al. 2015), as well as the shade tolerance of forages

grown in well-watered, enriched potting medium

under shade cloth inside greenhouses (Gaussoin

et al. 1988; Yang et al. 2012; Abraham et al. 2014).

Inside a greenhouse, the full sun treatment can be as

low as 49–67% of the ambient sunlight depending on

the infrastructure and covering (Wong et al. 1985a, b;

Abraham et al. 2014; Albaugh et al. 2014). Using

adequate irrigation, enriched potting medium, and

shade cloth, Lin et al. (1999) evaluated yield responses

of 30 forages grown within a hoop house under

moderate, dense, and non-shade, and Semchenko et al.

(2012) assessed the biomass production of 46 Estonian

grassland species grown under 10, 25, 50, and 100% of

ambient light.

There does not appear to be a clear, concise

definition of shade tolerance or how to measure it

when trying to rank plant species for shade tolerance.

Beard (1965) defined shade tolerance of turf grasses as

the ability to survive and maintain quality. When

ranking multiple species, some species inherently

yield more biomass per plant or unit of occupied area

than others under certain shade levels, thus methods

based on a species’ relative ability to tolerate shade are

required. Wong et al. (1985a) ranked the shade

tolerance of 12 tropical grass forages grown under

each of four shade levels based on the average scores

of several agronomic characteristics. Van Sambeek

et al. (2007) ranked 45 forages based on their

percentage change in forage production under dense

and moderate shade compared to non-shade.

A more comprehensive definition of shade toler-

ance is the capacity of plants to adapt morphologically

and physiologically to maintain productivity under

different light environments (Valladares et al. 2006;

Valladares and Niinemets 2008). These authors pro-

posed calculating the relative distance plasticity index

(RDPI) for traits of plants when grown under at least

three different light environments. Using the RDPI

methodology, we can generate a single value for each

species within each screening trial, thus multiple

RDPIs estimated from multiple trials of different

species can be statistically tested for differences

among species (Valladares et al. 2006).

The objectives of our study were to: (1) assess the

effect of non-, moderate, and dense shade on the

annual cumulative yields of 43 forages grown with

adequate water and nutrients and without root com-

petition from other plants; (2) evaluate whether RDPI

is a useful measure for quantifying shade tolerance.

Materials and methods

Study site

In spring 1996, an outdoor Shade Tolerance Screening

Laboratory (STSL) was constructed at the Horticulture

and Agroforestry Research Center (HARC) in New

Franklin, Missouri (92o 460W, 39o 010N). The STSL

sits on a 0.4 ha pad of limestone gravel laid over

permeable weed barrier for weed control and drainage.

A total of 72 posts (2.5 m tall) were set on a

4.9 9 4.9 m grid and connected by high tensile wires.

Shade cloth made of black polypropylene fabric was

hung over the top of the posts and wires, and along the

sides 0.1 m above the gravel to create nine structures in

a 3 9 3 array (Fig. O1 in Online Resource 1). Each

structure is 14.6 m long (north to south) and 4.9 m

wide, and 4.9 m away from other neighboring struc-

tures. Within each north–south block, three structures

Agroforest Syst

123

were randomly assigned to dense shade (80% shade

cloth, i.e., 20% of full sunlight), moderate shade (55%

shade cloth, 45% of full sunlight), and a non-shade

control (100% of full sunlight). Valves on pro-

grammable timers controlled duration of daily drip

irrigation to each pot. Each structure can hold 20 rows

of six pots on a 0.7 m 9 0.8 m spacing. Pots were

filled with a well-drained potting medium consisting of

composted pine bark, sphagnum peat moss, vermi-

culite, perlite, and sand (15:2:2:5:2 by volume), and

supplemented with slow-release nitrogen (7.9 g N per

10-L pot), micronutrient fertilizer, and a wetting agent.

During the growing season (May through September)

from 1996 to 2001, the study site had an average daily

temperature ranging from 16.6 �C to 27.7 �C. The

minimum and maximum temperature inside the shade

structures were recorded every 4 days from June to

September in 2001. The average minimum tempera-

tures within the non-, moderate, and dense shade

structures were 17.6, 19.2, and 22.2 �C, respectively,

and the average maximums were 35.2, 34.5, and

33.3 �C, respectively. Monthly precipitation during the

same time period averaged 10.7 cm and ranged from

1.9 to 21.7 cm (http://www.ncdc.noaa.gov/cdo-web/

datasets/GHCND/stations/GHCND:USC00236012/

detail). The ambient photosynthetic active radiation

(PAR) in an open field during the growing season

ranged from 36 to 44 mol m-2 day-1 from June to

September (unpublished data).

Plant material

Forages to be evaluated in the newly constructed

STSL were initially selected to replicate the species

screened by Lin et al. (1999). In subsequent years,

forages were selected to repeat screenings of certain

species, and new forages that were part of other on-

going agroforestry studies were also added. Seeds for

our screening trials were obtained from a local farm

supplier, the USDA Elsberry Plant Materials Center,

or harvested from native savanna and prairies in

Missouri. A total of 43 forages (39 species including 2

species having 2 cultivars each and 1 species with 3

cultivars) were chosen for our study (Table 1). These

forages contain annual and perennial, both cool-

season and warm-season, grasses and legumes that

were introduced or are native to the temperate region

of the United States based on information from the

USDA PLANTS database (USDA NRCS 2016).

Seed germination and seedling propagation

Seeds of grass species were germinated in seed starter

flats in a greenhouse in late March or early April each

year except 2001. Seeds of legumes were soaked in

85% rubbing alcohol for 1 min, and rinsed with

deionized water, then placed in petri dishes lined with

wet filter paper. When grass seedlings had one or two

true leaves, or legume seeds had emerging radicles,

three germinants were transplanted to a 442 cm3 pot

filled with a soil medium (Scott’s Metro Mix) until we

had 54 pots for each forage. When the legume

seedlings were established, a commercial rhizobial

inoculum labelled for each species was watered into

the potting medium. Both grasses and legumes were

watered with 0.11 g/L of Peter’s 20-20-20 NPK in the

greenhouse until April or May when six pots of each

cultivar were randomly transplanted into one row of

six black plastic pots (10 L each) in every shade

structure. In 2000, plants after the fall harvest were

overwintered under white plastic film and polyfoam

(Nursery blanket, Hummert Nursery, St. Louis) and

moved back into the structures in spring 2001.

Annual forage yield determination

Forage to a 10-cm stubble height was harvested in

summer at the boot stage (grasses) or early flowering

(legumes) and again in September or October from

1996 through 2000. In 2001, forage was harvested

when the majority of plants within a species started to

boot or flower resulting in up to four harvests. Forage

was oven dried at 70 �C for 72 h to minimize nitrogen

loss. Summer and fall yield of each cultivar within a

structure were summed to determine annual forage

yield (g pot-1) for statistical analysis.

Experimental design and data analysis

Annual yields from 1996 through 1999 were analyzed

using PROC Mixed in SAS 9.4 (SAS Inc., Gary, NC)

for each forage (Eq. O1 in Online Resource 1). Each

analysis is a split plot design with a randomized

complete blocking arrangement for shade (the whole-

plot) and year as the split-plot factor. Shade, year,

shade*year are fixed effects, and block, block*shade,

block*shade*year (residuals) are random effects. A

diagonal covariance structure (type = vc) was

applied. The forages screened in 2000 and 2001 were

Agroforest Syst

123

Table 1 Descriptive characteristics for 43 forages screened from 1996 to 2001 for forage yield under three shade treatments at the

Horticulture and Agroforestry Research Center (HARC), New Franklin, Missouri

Common namea Scientific namea Typeb Originc Lifecycle Years tested

96 97 98 99 00 01

Italian ryegrass Lolium perenne L. subsp. multiflorum

(Lam.) Husnot

CSG I Annual H H H H

Cheatgrass Bromus tectorum L. CSG I Annual H

Clustered fescue Festuca paradoxa Desv. CSG N Perennial H H

Red fescue Festuca rubra L. subsp. rubra CSG N Perennial H H

Kentucky bluegrass Poa pratensis L. CSG Both Perennial H H H

Orchardgrass ‘Benchmark’ Dactylis glomerata L. CSG I Perennial H H H

Perennial ryegrass Lolium perenne L. subsp. perenne CSG I Perennial H H

Redtop Agrostis gigantea Roth CSG I Perennial H H H H

Reed canarygrass Phalaris arundinacea L. CSG N Perennial H H H H

Smooth brome Bromus inermis Leyss. CSG Both Perennial H H H

Timothy Phleum pratense L. CSG I Perennial H H H

Alfalfa ‘Nitro’ Medicago sativa L. CSL I Perennial H H

Alfalfa ‘Victoria’ Medicago sativa L. CSL I Perennial H H

Alsike clover Trifolium hybridum L. CSL I Perennial H H

Crimson clover Trifolium incarnatum L. CSL I Annual H

Crownvetch ‘Penngift’ Securigera varia (L.) Lassen CSL I Perennial H H

Kura clover Trifolium ambiguum M. Bieb. CSL I Perennial H H

Red clover Trifolium pratense L. CSL I Perennial H H H

Sainfoin Onobrychis viciifolia Scop. CSL I Perennial H H

Subterranean clover Trifolium subterraneum L. CSL I Annual H

White clover Trifolium repens L. CSL I Perennial H H H H

Atra paspalum ‘Suerte’ Paspalum atratum Swallen WSG I Perennial H H

Bahiagrass ‘Argentine’ Paspalum notatum Fluegge WSG Both Perennial H H

Bahiagrass ‘Pensacola’ Paspalum notatum Fluegge WSG Both Perennial H H H

Bahiagrass ‘Tifton-9’ Paspalum notatum Fluegge WSG Both Perennial H H H

Bermudagrass Cynodon dactylon (L.) Pers. WSG I Perennial H H H

Eastern gamagrass Tripsacum dactyloides (L.) L. WSG N Perennial H H

Prairie cordgrass Spartina pectinata Bosc ex Link WSG N Perennial H H

Prairie dropseed Sporobolus heterolepis (A. Gray) A.

Gray

WSG N Perennial H H

Rhodes grass ‘Callide’ Chloris gayana Kunth WSG I Perennial H H

Switchgrass ‘Cave-in-rock’ Panicum virgatum L. WSG N Perennial H H H

Bird’s-foot trefoil ‘Norcen’ Lotus corniculatus L. WSL I Perennial H H

Bird’s-foot trefoil

‘rhizomatous’

Lotus corniculatus L. WSL I Perennial H H

Hoary ticktrefoil Desmodium canescens (L.) DC. WSL N Perennial H H H

Illinois bundleflower Desmanthus illinoensis (Michx.)

MacMill. ex B.L. Rob. & Fernald

WSL N Perennial H H H H

Korean clover Kummerowia stipulacea (Maxim.)

Makino

WSL I Annual H

Japanese clover Kummerowia striata (Thunb.) Schindl. WSL I Annual H H

Panicledleaf ticktrefoil Desmodium paniculatum (L.) DC. WSL N Perennial H H H

Purple prairie clover Dalea purpurea Vent. WSL N Perennial H

Agroforest Syst

123

analyzed using an additional autocorrelated covari-

ance structure—AR (1) to address the potential

correlation between first year and the second year

forage yields of the same plants.

Annual yield was also analyzed across years by

grouping forages as the following types: warm season

grasses (WSG, C4), cool season grasses (CSG, C3),

warm season legumes (WSL, C3), or cool season

legumes (CSL, C3), using Eq. O2 (Online Resource 1).

Shade, type, shade*type are fixed effects, species

(type), i.e., species nested in type, and the residuals are

random effects. Annual forage yields were natural-log

transformed before analysis, and then back-trans-

formed for data presentation. Tukey method

(a = 0.05) was applied for the mean separations.

RDPIs were calculated according to Valladares

et al. (2006) as the pairwise comparisons across

replications of annual forage yield under non-, mod-

erate, and dense shade:

RDPI ¼P Xjm�Xj0m0j j

XjmþXj0m0

�

n ð1Þ

where Xjm or Xj0m0 is the forage yield; m or m0 is the

mth block (m or m0 = 1, 2, 3, but m = m0) subjected

to light treatment j or j0 (j or j0 = 1, 2, 3, but j = j0),and n is the total number of all possible pairs of block

and shade treatment in a given year. Missing yield

values of each forage were imputed as the average of

the other two replicates of the same treatment, thus

n = 27. RDPIs were calculated for each cultivar in

each year it was screened. For multiple-cultivar

species, cultivars were treated as replicates, and the

RDPI estimated for that species is the average across

all cultivars and years. To include more forages and

increase the accuracy of estimation, RDPIs were also

calculated from the original data for 27 forages (22

species with 5 species having 2 cultivars each)

screened by Lin et al. (1999) under the same shade

treatments as in this study. RDPIs were natural-log

transformed and compared both among species and

forb types in PROC MIXED, and Tukey method

(a = 0.05) was used for the mean separations.

Results

Annual forage yield

Annual forage yield differed among shade treatments

for 30 of 43 forages (p\ 0.05, Table 2). Year effect

was also significant (p \ 0.05, Table 2) and differ-

ences in yield among years existed for 31 out of the 35

forages screened for more than 1 year (Table O1 in

Online Resource 1). For forages that were tested

between 1996 and 1999, average forage yield across

all shade treatments was generally highest in 1997

(except Rhodes grass ‘Callide’ had higher yield in

1996 than in 1997). No differences in yield were found

among the same forages tested in 1996 and 1998

except Illinois bundle flower with higher yield in 1996

than 1998. Forage yields were equal for 10 forages in

1998 and 1999 with another nine forages yielding less

in 1999 than in 1998. In the two-consecutive-growing-

season trials (2000, 2001), annual forage yields in

2001 of overwintered plants across all shade treat-

ments were higher in 11 out of 13 forages than the

yields of plants started as seedlings in 2000. For the

remaining two forages, alfalfa ‘Nitro’ had equal

Table 1 continued

Common namea Scientific namea Typeb Originc Lifecycle Years tested

96 97 98 99 00 01

Roundhead lespedeza Lespedeza capitata Michx. WSL N Perennial H

Showy ticktrefoil Desmodium canadense (L.) DC. WSL N Perennial H

Slender lespedeza Lespedeza virginica (L.) Britton WSL N Perennial H

Strawberry clover Trifolium fragiferum L. WSL Both Perennial H H H H

a Both common and scientific names followed the information in USDA PLANTS database (USDA NRCS 2016)b CSG cool season grasses (C3), WSG warm season grasses (C4), CSL cool season legumes, WSL warm season legumesc N native to US, I introduced to US

Agroforest Syst

123

Table

2P

rob

abil

ity

of

sig

nifi

can

tF

-val

ue

of

the

fix

edef

fect

sin

the

mix

edef

fect

sm

od

els

for

each

fora

ge,

the

aver

age

ann

ual

fora

ge

yie

ld(g

po

t-1),

and

95

%co

nfi

den

cein

terv

al

(CI)

for

43

fora

ges

gro

wn

un

der

no

n-,

mo

der

ate,

and

den

sesh

ade

bet

wee

n1

99

6an

d2

00

1

Fo

rag

esN

um

ber

of

yea

rste

sted

aF

bN

on

-sh

ade

(fu

llsu

n)

Mo

der

ate

shad

e

(45

%o

ffu

llsu

n)

Den

sesh

ade

(20

%o

ffu

llsu

n)

Sh

ade

(S)

S*

YY

ear

(Y)

Bio

mas

sc

(gp

ot-

1)

95

%C

IB

iom

ass

(gp

ot-

1)

95

%C

IB

iom

ass

(gp

ot-

1)

95

%C

I

Gro

up

1:

45

%=

10

0%

=2

0%

Clu

ster

edfe

scu

e2

,C1

.05

NS

0.0

5N

S1

93

.71

**

*2

2.6

a1

8.2

–2

8.1

24

.9a

20

.1–

30

.92

0.7

a1

6.7

–2

5.7

Eas

tern

gam

agra

ss2

,C1

.15

NS

0.8

7N

S6

8.2

3*

**

26

.6a

18

.3–

38

.73

8.2

a2

6.3

–5

5.6

31

.5a

21

.7–

45

.8

Ch

eatg

rass

11

.63

NS

––

16

.2a

11

.0–

23

.71

5.8

a1

0.8

–2

3.1

11

.0a

7.5

–1

6.1

Ku

racl

ov

er2

,C2

.24

NS

4.5

0N

S1

51

.24

**

*8

.2a

5.3

–1

2.6

7.9

a5

.2–

12

.25

.6a

3.6

–8

.6

Sle

nd

erle

sped

eza

12

.29

NS

––

8.4

a5

.2–

13

.61

1.8

a7

.3–

19

.16

.9a

4.3

–1

1.1

Sh

ow

yti

cktr

efo

il1

3.2

5N

S–

–2

7.4

a1

9.5

–3

8.4

46

.8a

33

.3–

65

.73

3.7

a2

4.0

–4

7.4

Atr

ap

asp

alu

m‘S

uer

te’

23

.32

NS

1.0

0N

S2

3.2

9*

*1

20

.9a

10

5.4

–1

38

.61

36

.2a

11

8.8

–1

56

.21

48

.3a

12

9.4

–1

70

.1

Ro

un

dh

ead

lesp

edez

a1

3.7

5N

S–

–6

.8a

5.2

–9

.09

.9a

7.5

–1

3.1

6.9

a5

.2–

9.1

Sai

nfo

in2

,C3

.94

NS

5.0

7N

S1

21

.10

**

*2

0.0

a1

2.6

–3

1.8

16

.2a

10

.2–

25

.89

.0a

5.7

–1

4.3

Per

enn

ial

ryeg

rass

2,C

4.6

3N

S1

.10

NS

55

.49

**

*2

3.1

a1

4.7

–3

6.4

21

.5a

13

.6–

33

.81

8.3

a1

1.6

–2

8.9

Red

top

45

.57

NS

2.3

4N

S1

7.6

4*

**

43

.2a

34

.8–

53

.65

9.2

a4

7.7

–7

3.4

38

.2a

30

.8–

47

.3

Orc

har

dg

rass

‘Ben

chm

ark

’

35

.83

NS

0.6

7N

S2

7.6

1*

**

34

.1a

30

.4–

38

.34

3.4

a3

8.7

–4

8.7

36

.6a

32

.6–

41

.1

Su

bte

rran

ean

clo

ver

16

.40

NS

––

7.1

a2

.5–

20

.21

1.6

a4

.0–

33

.03

.4a

1.2

–9

.7

Gro

up

2:

45

%[

20

%;

10

0%[

20

%

Bir

d’s

-fo

ot

tref

oil

‘rh

izo

mat

ou

s’

2,C

12

.32

*0

.09

NS

17

4.8

8*

**

44

.6a

33

.5–

59

.35

1.2

a3

8.5

–6

8.2

22

.1b

16

.6–

29

.4

Ko

rean

clo

ver

11

2.1

9*

––

47

.1a

30

.0–

73

.93

7.6

a2

4.0

–5

8.9

16

.2b

10

.3–

25

.4

Illi

no

isb

un

dle

flo

wer

41

4.5

1*

0.9

3N

S1

0.8

0*

**

33

.7a

26

.9–

42

.24

1.4

a3

3.1

–5

1.8

20

.1b

16

.1–

25

.2

Ber

mu

dag

rass

31

7.1

0*

1.4

7N

S3

.08

NS

12

5.3

a1

01

.9–

15

4.1

14

0.1

a1

13

.9–

17

2.4

78

.6b

63

.9–

96

.7

Als

ike

clo

ver

2,C

25

.56

**

6.3

1*

15

.42

**

24

.1a

18

.0–

32

.43

0.2

a2

2.5

–4

0.4

11

.3b

8.4

–1

5.2

Rh

od

esg

rass

‘Cal

lid

e’2

86

.72

**

*1

.06

NS

58

.53

**

*1

01

a8

1.4

–1

25

.39

1.1

a7

3.5

–1

13

57

.1b

46

.0–

70

.8

Jap

anes

ecl

ov

er2

33

.19

**

1.1

8N

S3

4.7

1*

*4

1.2

a3

0.4

–5

5.8

50

.3a

37

.1–

68

.11

3.4

b9

.9–

18

.2

Pu

rple

pra

irie

clo

ver

13

5.2

5*

*–

–8

.0a

5.0

–1

2.7

6.3

a3

.9–

10

.00

.9b

0.6

–1

.4

Red

fesc

ue

2,C

54

.82

**

26

.00

**

11

.01

*3

5.1

a2

6.5

–4

6.3

34

.5a

26

.1–

45

.71

2.7

b9

.6–

16

.8

Pra

irie

cord

gra

ss2

57

.18

**

3.4

9N

S0

.08

NS

35

.1a

24

.8–

49

.83

5.0

a2

4.7

–4

9.6

11

.9b

8.4

–1

6.9

Agroforest Syst

123

Table

2co

nti

nu

ed

Fo

rag

esN

um

ber

of

yea

rste

sted

aF

bN

on

-sh

ade

(fu

llsu

n)

Mo

der

ate

shad

e

(45

%o

ffu

llsu

n)

Den

sesh

ade

(20

%o

ffu

llsu

n)

Sh

ade

(S)

S*

YY

ear

(Y)

Bio

mas

sc

(gp

ot-

1)

95

%C

IB

iom

ass

(gp

ot-

1)

95

%C

IB

iom

ass

(gp

ot-

1)

95

%C

I

Sw

itch

gra

ss‘C

ave-

in-

Ro

ck’

35

7.4

1*

*1

.13

NS

11

.90

**

89

.2a

77

.4–

10

2.8

85

.9a

74

.5–

99

.03

8.5

b3

3.4

–4

4.4

Pra

irie

dro

pse

ed2

,C7

7.8

7*

**

0.0

1N

S4

0.1

6*

**

37

.3a

27

.8–

49

.92

6.0

a1

9.4

–3

4.8

6.0

b4

.5–

8.0

Gro

up

3:

45

%=

10

0%

,1

00

%=

20

%,

45

%[

20

%

Alf

alfa

‘Nit

ro’

2,C

9.0

6*

0.2

4N

S5

.42

NS

45

.2ab

29

.4–

69

.74

8.7

a3

1.6

–7

5.1

18

.5b

12

.0–

28

.5

Alf

alfa

‘Vic

tori

a’2

,C7

.41

*0

.71

NS

13

3.7

8*

**

43

.7ab

30

.2–

63

.15

4.1

a3

7.4

–7

8.1

22

.7b

15

.7–

32

.8

Bir

d’s

-fo

ot

tref

oil

‘No

rcen

’

2,C

7.5

5*

1.4

7N

S1

11

.07

**

*4

1.8

ab2

9.3

–5

9.6

57

.9a

40

.6–

82

.62

4.2

b1

7.0

–3

4.5

Bah

iag

rass

‘Arg

enti

ne’

21

8.8

5*

*1

.22

NS

12

.66

*6

1.8

ab5

3.9

–7

0.9

77

.8a

67

.8–

89

.24

9.7

b4

3.3

–5

7.0

Bah

iag

rass

‘Pen

saco

la’

37

.05

*2

.45

NS

11

.14

**

69

.6ab

55

.1–

87

.88

2.3

a6

5.2

–1

03

.94

8.0

b3

8.0

–6

0.5

Bah

iag

rass

‘Tif

ton

-9’

31

3.7

0*

1.1

5N

S2

.33

NS

71

.5ab

60

.8–

84

.18

6.2

a7

3.3

–1

01

.35

0.8

b4

3.2

–5

9.8

Cri

mso

ncl

ov

er1

10

.56

*–

–1

3.0

ab4

.8–

35

.02

6.5

a9

.8–

71

.24

.3b

1.6

–1

1.5

Red

clo

ver

31

3.6

4*

1.3

8N

S2

1.1

2*

**

36

.7ab

26

.5–

50

.76

2.1

a4

4.9

–8

5.9

21

.3b

15

.4–

29

.4

Tim

oth

y3

8.8

9*

0.8

2N

S2

2.7

3*

**

29

.6ab

20

.3–

43

.34

1.4

a2

8.3

–6

0.6

18

.7b

12

.8–

27

.4

Gro

up

4:

45

%[

10

0%

,1

00

%=

20

%,

45

%[

20

%

Sm

oo

thb

rom

e3

11

.35

*1

.48

NS

68

.26

**

*3

0.6

b2

7.6

–3

3.9

40

.4a

36

.4–

44

.83

1.2

b2

8.1

–3

4.6

Pan

icle

dle

afti

cktr

efo

il3

12

.26

*0

.66

NS

8.0

9*

*4

4.8

b3

5.7

–5

6.1

75

.0a

59

.8–

94

.04

5.4

b3

6.2

–5

6.8

Ital

ian

ryeg

rass

41

7.3

1*

8.8

3*

**

49

.74

**

*3

5.1

b3

0.6

–4

0.3

54

.1a

47

.1–

62

.03

5.7

b3

1.1

–4

0.9

Ken

tuck

yb

lueg

rass

32

3.8

5*

*1

7.7

3*

**

20

3.7

1*

**

19

.1b

16

.9–

21

.63

1.4

a2

7.8

–3

5.5

20

.7b

18

.3–

23

.4

Wh

ite

clo

ver

42

6.5

9*

*4

.55

**

40

.82

**

*2

8.9

b2

2.1

–3

7.8

53

.6a

41

.0–

70

.12

7.3

b2

0.9

–3

5.7

Str

awb

erry

clo

ver

43

1.5

1*

*2

.49

NS

33

.92

**

*2

4.4

b1

8.8

–3

1.6

58

.6a

45

.1–

75

.92

3.1

b1

7.8

–3

0.0

Gro

up

5:

45

%=

20

%,

45

%[

10

0%

,2

0%[

or=

10

0%

Ree

dca

nar

yg

rass

47

.03

*1

.80

NS

8.4

1*

*4

0.5

b3

4.4

–4

7.6

57

.7a

49

.0–

67

.94

7.9

ab4

0.7

–5

6.4

Ho

ary

tick

tref

oil

31

8.8

7*

*1

.51

NS

41

.51

**

*2

5.9

b2

1.7

–3

1.0

48

.6a

40

.6–

58

.13

6.5

ab3

0.5

–4

3.6

Cro

wn

vet

ch‘P

enn

gif

t’2

,C1

3.8

4*

1.5

6N

S1

09

.85

**

*2

8.3

b2

2.9

–3

5.0

37

.9a

30

.7–

46

.83

7.8

a3

0.6

–4

6.7

aC

:co

nse

cuti

vel

yg

row

nfr

om

20

00

to2

00

1b

NS

:n

ot

sig

nifi

can

t(p

C0

.05

);*

0.0

1B

p\

0.0

5;

**

0.0

01B

p\

0.0

1;

***

p\

0.0

00

1c

Mea

ns

wit

hsa

me

des

ign

atio

nle

tter

sw

ith

ina

row

are

no

tsi

gn

ifica

ntl

yd

iffe

ren

t(a

=0

.05

)b

yT

uk

ey’s

test

.V

alu

esw

ere

bac

ktr

ansf

orm

edfr

om

nat

ura

llo

gar

ith

mto

ori

gin

al

scal

e

Agroforest Syst

123

biomass in 2000 and 2001, and alsike clover had

higher biomass in 2000 than in 2001.

Among the 43 forages, five forages showed an

interaction between shade and year (p\ 0.05,

Table 2). Mostly, the interaction occurred due to more

variable yields in the non-shade control (Fig. 1) that

may have been in response to ineffective irrigation

during hot dry weather in some years. Although pots

were automatically irrigated, evidence suggests the soil

surface may have formed a crust so that irrigation water

flowed over the surface without adequately wetting the

soil. Light, long duration precipitation usually elimi-

nated these crusts. Overall, forage yields were the least

during the growing season (May to September) in

1999, which also had the least precipitation: 16 days of

C5 mm and 11 days of C10 mm (Fig. 2) compared to

the average 24.5 and 17.7 days, respectively. The 1999

season also had the second highest number of days

(70 days) with a maximum temperature C30 �C(Fig. 2). The 1998 growing season had the most days

(75) with a maximum temperatureC30 �C; however, it

had more frequent precipitation with 29 days of

C5 mm and 21 days of C10 mm (Fig. 2).

The shade responses of forage yield showed five

major patterns (Table 2). Forages in group 1 showed

no yield differences under all three shade treatments,

indicating these species have a high degree of shade

tolerance. This group included cheat grass, slender

lespedeza, showy ticktrefoil, roundhead lespedeza,

subterranean clover, clustered fescue, eastern gama-

grass, Kura clover, atra paspalum, sainfoin, perennial

ryegrass, redtop, and orchardgrass.

Fig. 1 Annual biomass yield (g pot-1) responses to non-, moderate, and dense shade of the five forages among the 43 forages with

significant interactions (p\ 0.05) between shade and year from 1996 to 2001

Agroforest Syst

123

The forages in group 2 had equal yields between

moderate shade and the non-shade control, but yield

under dense shade averaged 40% of that under

moderate shade and the control. This suggests group

2 is tolerant to moderate, but not dense shade. This

group included bird’s-foot trefoil ‘rhizomatous’, Illi-

nois bundleflower, Bermudagrass, alsike clover,

Rhodes grass, Japanese clover, red fescue, prairie

cordgrass, switchgrass, purple prairie clover, prairie

drop seed, and Korean clover.

In group 3, forage yield was greater under moderate

than under dense shade with no differences between

the control and dense shade nor between the control

and moderate shade. This group, of which the yield

under dense shade averaged 44% of that under

moderate shade, included crimson clover, the two

alfalfa cultivars, bird’s-foot trefoil ‘Norcen’, all 3

bahiagrass cultivars, red clover, and timothy.

In group 4, forages grown under moderate shade

had higher yield than the control or dense shade with

no differences between the latter two treatments.

Yields under the control and dense shade averaged

about 59 and 60% of that under moderate shade,

respectively. Forages in this group included white

clover, strawberry clover, Italian ryegrass, Kentucky

bluegrass, smooth brome, and panicledleaf ticktrefoil.

For group 5, forages under moderate shade grew

better than the control, while plants under dense shade

had biomass growth equal to or exceeding that of the

control. The yield of the forages in the control was

66%, on average, of that under moderate shade, and

77% of dense shade. Forages in this group included

crownvetch, reed canary grass, and hoary ticktrefoil.

A shade effect was detected when grouping forages

by forb types (p\ 0.0001, Table O2 in Online

Resource 1). Although an interaction existed between

forb types and shade treatments (p\ 0.0001), all four

forb types performed best under moderate shade,

followed by the control and dense shade (Fig. 3).

Grasses generally yielded higher biomass than

legumes. Average forage yields declined from warm

season grasses, cool season grasses, and then to warm

season legumes and the cool season legumes as the

two least productive types (p = 0.018).

Relative distance plasticity index

Relative distance plasticity indices for forage yield

ranged from 0.104 to 0.567 for 45 species (Table 3)

with differences existing among species (p\ 0.0001).

The 13 species with the lowest RDPIs and presumably

the most shade tolerant in terms of forage production

Fig. 2 Number of days

with maximum air

temperature C25 or C30 �Cand number of days with

rainfall C5 or C10 mm from

May to September in

1996–2001 in New Franklin,

Missouri

Agroforest Syst

123

were clustered fescue, atra paspalum, crownvetch,

smooth brome, cheatgrass, orchardgrass, roundhead

lespedeza, reed canarygrass, hoary ticktrefoil, bahia-

grass, eastern gamagrass, panicledleaf ticktrefoil, and

perennial ryegrass (Table 3).

Analysis of the four forage types showed that

grasses have lower RDPI than legumes (p\ 0.0001),

with the CSG (C3) having the lowest average RDPI

followed by the WSG (C4), then the WSL (C3), and,

finally, the CSL (C3) having the highest average RDPI

(Table 3).

Discussion

Shade effects on forage yield

Compared to the non-shade control, 34 of the 43

forages showed no differences and nine forages had

greater yields when grown under moderate shade

(Table 2). Likewise, 30 forages showed no differences

and one forage had greater yield than the control when

grown under dense shade (Table 2). In a similar study

by Lin et al. (1999), 13 forages (out of 27) in the two

summer-fall trials, and nine forages in the spring–

summer trial produced no differences between mod-

erate shade and the control, and three forages in the

summer-fall trials had higher yields under moderate

shade than the control. Three forages in the summer-

fall trials and five forages in the spring–summer trial

showed no differences in forage yield between dense

shade and the control. Forages that performed equally

well under moderate shade and the control from both

studies included smooth brome, hoary ticktrefoil,

panicledleaf ticktrefoil, Kentucky bluegrass, orchard-

grass, perennial ryegrass, timothy, and alfalfa ‘Victo-

ria’. Semchenko et al. (2012) discovered that 43 out of

46 temperate grassland species in Estonia grew more

biomass under moderate shade (50% of full sunlight)

than non-shade. These Estonian species include many

grasses, among which some are forages or closely

related to the forages tested in our study, such as soft

brome (Bromus hordeaceus), sweet vernalgrass (An-

thoxanthum odoratum), Kentucky bluegrass, colonial

bentgrass (Agrostis capillaris), creeping bentgrass

(Agrostis stolonifera), orchardgrass, red fescue, tall

fescue (Schedonorus arundinaceus, syn. Festuca

arundinacea in their original text), timothy, and reed

canarygrass.

Similar shade response patterns, i.e., biomass either

plateaued or peaked at moderate shade, have been

found with seedlings of temperate tree species. Loach

(1970) reported that seedlings of five species, both

shade tolerant (Fagus grandifolia and Acer rubrum)

and intolerant (Liriodendron tulipifera, Quercus

rubra, and Populus tremuloides), had comparable or

higher root, stem, and leaf biomass under moderate

shade (44% of full sun) than the control; but biomasses

were lowest under dense shade (17 and 3% of full sun).

Other studies, however, have reported shade

responses that do not plateau or peak at moderate

shade. Devkota (1997) reported a linear increase in

shoot dry weight of ten pasture species across five

shade levels from 14 to 78% of ambient PAR in a

greenhouse study. A similar linear increase was found

in biomass of shade-treated orchardgrass from 10, 30,

to 100% of full sun inside a greenhouse (Abraham

et al. 2014). Both aboveground and belowground

biomass of switchgrass ‘Alamo’ increased when

grown inside a greenhouse as light intensity increased

from 11, to 23, to 31, to 49%, and to 100% of full sun

(Albaugh et al. 2014). Jiang et al. (2004) found

biomass of seashore paspalum (Paspalum vaginatum

Swartz) and hybrid Bermudagrass (Cynodon dactylon

L. 9 C. transvaalensis Burtt Davy) cultivars

increased from 10 to 30%, and then to 100% of full

sun.

Fig. 3 Annual biomass yield (g pot-1) responses to non-,

moderate, and dense shade between 1996 and 2001 by forage

types: CSG cool season grasses, WSG warm season grasses,

WSL warm season legumes, and CSL cool season legumes

Agroforest Syst

123

Table 3 Relative distance plasticity index (RDPI), 95% confidence interval (CI), and rank for annual forage yield (g pot-1) of 45

forage species tested between 1994 and 2001

Species Groupa RDPIb 95% CI Rank Number of cultivarsc

1996–2001 1994–1995

Clustered fescue CSG 0.104df 0.063–0.171 1 1 0

Atra paspalum WSG 0.128abcd 0.078–0.210 2 1 0

Crownvetch CSL 0.134abcd 0.081–0.220 3 1 0

Smooth brome CSG 0.136d 0.102–0.181 4 1* 1

Cheatgrass CSG 0.145abcd 0.088–0.238 5 1 0

Orchardgrass CSG 0.148d 0.117–0.187 6 1* 2

Roundhead lespedeza WSL 0.160abcd 0.079–0.323 7 1 0

Reed canarygrass CSG 0.161bcd 0.114–0.229 8 1 0

Hoary ticktrefoil WSL 0.162cde 0.122–0.216 9 1* 1

Bahiagrass WSG 0.163cd 0.127–0.208 10 3 0

Eastern gamagrass WSG 0.172abcd 0.104–0.282 11 1 0

Panicledleaf ticktrefoil WSL 0.174bcd 0.131–0.232 12 1* 1

Perennial ryegrass CSG 0.175abcd 0.128–0.240 13 1* 1

Showy ticktrefoil WSL 0.189abcd 0.094–0.380 14 1 0

Italian ryegrass CSG 0.194abcd 0.137–0.276 15 1 0

Redtop CSG 0.197abcd 0.139–0.279 16 1 0

Kentucky bluegrass CSG 0.208abcd 0.156–0.277 17 1* 1

Tall fescue CSG 0.226abcd 0.170–0.301 18 0 2

Rhodes grass WSG 0.228abcd 0.139–0.374 19 1 0

Slender lespedeza WSL 0.247abcd 0.150–0.406 20 1* 1

Kura clover CSL 0.249abcd 0.152–0.409 21 1 0

Sericea lespedezad WSL 0.254abcd 0.155–0.418 22 0 1

Timothy CSG 0.257abcd 0.193–0.343 23 1* 1

Bermudagrass WSG 0.263abcd 0.185–0.373 24 1* 1

Illinois bundleflower WSL 0.267abcd 0.188–0.379 25 1 0

Sainfoin CSL 0.269abcd 0.164–0.443 26 1 0

White clover CSL 0.275abcd 0.211–0.359 27 1* 1

Alfalfa CSL 0.281abcd 0.225–0.351 28 2 2

Switchgrass WSG 0.322abcf 0.242–0.429 29 1* 1

Big bluestem WSG 0.322abcd 0.215–0.483 30 0 1

Prairie cordgrass WSG 0.33abcd 0.201–0.542 31 1 0

Japanese clover WSL 0.331abcf 0.242–0.453 32 1* 1

Bird’s–foot trefoil WSL 0.336abe 0.269–0.419 33 2* 2

Indiangrasse WSG 0.34abcd 0.226–0.509 34 0 1

Strawberry clover WSL 0.343abcd 0.242–0.487 35 1 0

Red clover CSL 0.382abe 0.269–0.477 36 1* 1

Alsike clover CSL 0.394abe 0.279–0.523 37 1* 1

Red fescue CSG 0.401abcd 0.240–0.647 38 1 0

Korean clover WSL 0.358a 0.307–0.523 39 1* 2

Crimson clover CSL 0.432abcd 0.263–0.710 40 1 0

Berseem cloverf CSL 0.459ab 0.306–0.689 41 0 1

Buffalograssg WSG 0.468abcd 0.232–0.945 42 0 1

Agroforest Syst

123

Plant responses to decreasing light can range from

linear decreases in forage yield to those that plateau or

peaked under moderate shade as found in Lin et al.

(1999), Semchenko et al. (2012), and our study.

Because net assimilation in C3 plants saturates around

50–60% of maximum sunlight, it is possible plants

grown under moderate shade achieved maximum net

assimilation on most days, while plants grown without

shade absorbed surplus light resulting in high rates of

dark respiration which reduced photosynthetic effi-

ciency. Maximum air temperatures in the non-shade

structures were 0.7 and 1.9 �C higher than under

moderate and dense shade which were covered with

black shade cloth, further increasing dark respiration

rates for plants exposed to full sun. Similar temper-

ature differences among shade structures compared

with the non-shade control was also noted by

Semchenko et al. (2012).

Interactions between shade and year occurred for

five species (Fig. 1) in part due to variable responses in

forage yield from year to year, especially in the non-

shade control. Lower biomass in 2000 was in part

because most species were planted late and harvested

only in the fall of 2000, while in 2001 most species had

2–4 harvests depending on harvest recovery time and

initiation of flowering. Of the 13 perennial forages

screened in 2000 and 2001, only red fescue and alsike

clover showed an interaction between shade treatment

and year, suggesting yield response to shade does not

change as most perennial forages mature from first-

year seedlings to established plants.

Relative distance plasticity index

Forage species that show the least differences in

biomass yield across non-, moderate, and dense shade

(group 1), or at least the latter two treatments (group

5), should in theory have the lowest RDPIs and ought

to be the most shade tolerant species. Among the top

13 species with the lowest RDPIs (Table 3), clustered

fescue, atra paspalum, cheatgrass, orchardgrass,

roundhead lespedeza, eastern gamagrass, and peren-

nial ryegrass are in group 1, while crownvetch, reed

canarygrass, and hoary ticktrefoil are in group 5.

Species with reduced biomass under dense shade

compared to moderate and/or non-shade (groups 2, 3,

and 4) should have higher RDPIs. None of the 12

species in group 2, only bahiagrass out of nine species

in group 3, and smooth brome and panicledleaf

ticktrefoil out of six species in group 4 have low

RDPIs. The calculation of RDPI of smooth brome

(0.136) and panicledleaf ticktrefoid (0.174) included

biomass data from Lin et al. (1999), which showed

smaller differences in biomass than our study across

all three shade levels.

RDPIs confirm relative shade tolerance of reported

observations for several forages. Orchardgrass is

reported to be productive under a wide range of light

Table 3 continued

Species Groupa RDPIb 95% CI Rank Number of cultivarsc

1996–2001 1994–1995

Subterranean clover CSL 0.494abc 0.301–0.811 43 1 0

Prairie dropseed WSG 0.494abc 0.301–0.812 44 1 0

Purple prairie clover WSL 0.567abcd 0.281–1.144 45 1 0

Forb types CSG 0.183B 0.163–0.206 I

WSG 0.252A 0.217–0.292 II

WSL 0.274A 0.242–0.311 III

CSL 0.314A 0.274–0.360 IV

* indicates the number of cultivars of that species shared with Lin et al. (1999)a CSG cool season grasses, WSG warm season grasses, CSL cool season legumes, WSL warm season legumesb RDPIs with the same lower case or upper case letters in the same column are not significantly different (a = 0.05). RDPIs and CIs

are all back transformed from natural log to original scalec Number of cultivars tested in each period of timed–f Forages that are from Lin et al. (1999) but their scientific names do not appear in the text other than this table. sericea lespedeza:

Lespedeza cuneate; Indiangrass: Sorghastrum nutans; berseem clover: Trifolium alexandrinum; buffalograss: Bouteloua dactyloides

Agroforest Syst

123

conditions and is recognized as a shade tolerant

species (Blake et al. 1966; Christie and McElroy

1995). Our RDPI calculated as the average of two

cultivars across several trials (six trials for ‘Bench-

mark’ and three for ‘Justus’) was 0.148, which ranked

orchardgrass as the 6th most shade tolerant out of the

45 species examined. Atra paspalum, a shade tolerant

C4 grass recommended for tropical agroforestry

systems had a low RDPI (0.128), placing it the 2nd

out of 45 species. White clover, reported to be shade

intolerant by Christie and McElroy (1995), has a

relatively high RDPI (0.275) and ranked 27th on the

list of 45 species.

We also assessed the usefulness of RDPI to rank

forage species for shade tolerance by comparing

published relative ranks to our estimated RDPIs.

RDPIs align well with the rankings by Feldhake and

Belesky (2009) for orchardgrass (0.148) being more

shade tolerant than tall fescue (0.226). RDPI and

rankings do not align so well for the shade tolerance

rankings reported by Beard (1965) for red fescue

(0.394) [ perennial ryegrass (0.175) [ tall fescue

(0.226)[Kentucky bluegrass (0.208), or by Kephart

et al. (1992) for tall fescue (0.226)[reed canarygrass

(0.161) [ switchgrass (0.322). RDPIs support the

rankings by Devkota et al. (1997) of orchardgrass

(0.148) being more shade tolerant than perennial

ryegrass (0.175) but not for subterranean clover

(0.494) being more shade tolerant than white clover

(0.275). Discrepancies among relative rankings may

be in part due to measuring different growth responses,

i.e., survival and plant health; amount of root compe-

tition (pasture compared to a few plants in a large pot);

or study duration (a couple months compared to

multiple harvests over an entire growing season). Van

Sambeek et al. (2007) investigated species with widely

differing shade tolerance from multiple screening

trials by comparing the relative changes in forage

yields from plants under moderate and dense shade to

those under non-shade, and then ranked them on a

scale between 0 (least tolerant) and 100% (most

tolerant). There is reasonable agreement between the

reported percentile ranks and RDPIs of our study

except for bahiagrass, Bermudagrass, switchgrass, and

red clover.

C3 and C4 species generally have different degrees

of shade tolerance. Kephart et al. (1992) found that in

response to reduced irradiance, C3 grasses like tall

fescue, reed canarygrass, and deertongue grass

(Dichanthelium clandestinum) have smaller relative

reductions in biomass than C4 grasses such as

switchgrass and big bluestem (Andropogon gerardii),

indicating more shade tolerance within C3 grasses

than C4. This is in agreement with our finding that C3

grasses (cool season) are more resilient to shade

(lower RDPI) in forage yield than C4 grasses (warm

season). However, there was no difference for RDPIs

between legumes (all C3) and C4 grasses.

Conclusion

Multiple screenings evaluating responses of forages to

moderate and dense shade showed most forages did

not change or increased in forage yield, when grown

with adequate water and nutrients in the absence of

root competition from other plants. RDPI that mea-

sures the resilience of a plant to adapt when grown

along an environmental gradient has acceptable con-

cordance with known shade tolerance rankings. Low

RDPIs indicate small changes in yield for forage

species that can adapt to moderate and dense shade.

Several agroforestry systems are characterized by

continuums from moderate to dense shade in the

presence of trees. Our results indicate that for most

forages, yield can be maintained or improved under

moderate shade compared to open grown forages

when tree root competition is minimized. RDPI can be

a useful and convenient indicator for shade tolerance.

Agroforestry practitioners may want to choose forages

with low RDPIs, because these species are more likely

to maintain biomass yield in their agroforestry prac-

tices as trees grow and canopies close.

Acknowledgements This work was funded by the University

of Missouri Center for Agroforestry under cooperative

agreement AG-02100251 with the USDA-ARS Dale Bumpers

Small Farm Research Center, Booneville, AR, and CR

826704-01-0 with the US EPA. We want to express our

appreciation to the HARC staff, especially Steve Kirk for

maintaining the studies in the STSL and harvesting the forages,

Robert McGraw, our retired agronomist, for providing seeds and

guidance, and John Stanovick, a USDA Forest Service

statistician, for helping us on statistical analysis.

References

Abraham EM, Kyriazopoulos AP, Parissi ZM, Kostopoulou P,

Karatassiou M, Anjalanidou K, Katsouta C (2014) Growth,

Agroforest Syst

123

dry matter production, phenotypic plasticity, and nutritive

value of three natural populations of Dactylis glomerata L.

under various shading treatments. Agrofor Syst

88:287–299. doi:10.1007/s10457-014-9682-9

Albaugh JM, Albaugh TJ, Heiderman RR, Leggett Z, Stape JL,

King K, O’Neill KP, King JS (2014) Evaluating changes in

switchgrass physiology, biomass, and light-use efficiency

under artificial shade to estimate yields if intercropped with

Pinus taeda L. Agrofor Syst 88:489–503. doi:10.1007/

s10457-014-9708-3

Beard JB (1965) Factors in the adaption of turfgrass to shade.

Agron J 57:457. doi:10.2134/agronj1965.0002196200570

0050015x

Blake CT, Chamblee DS, Woodhouse WW (1966) Influence of

some environmental and management factors on the per-

sistence of Ladino clover in association with orchardgrass.

Agron J 58:487–489. doi:10.2134/agronj1966.00021962

005800050009x

Callaway R (2007) Positive interactions and interdependence in

plant communities. Springer, Dordrecht

Christie BR, McElroy AR (1995) Orchardgrass. In: Barnes RF,

Miller DA, Nelson CJ (eds) Forages, 5th edn. Iowa State

University Press, Ames, pp 325–334

Devkota NR, Kemp PD, Hodgson J (1997) Screening pasture

species for shade tolerance. Proc Agron Soc NZ

27:119–128

Ehret M, Graß R, Wachendorf M (2015) The effect of shade and

shade material on white clover/perennial ryegrass mixtures

for temperate agroforestry systems. Agrofor Syst

89:557–570. doi:10.1007/s10457-015-9791-0

Feldhake CM, Belesky DP (2009) Photosynthetically active

radiation use efficiency of Dactylis glomerata and Sche-

donorus phoenix along a hardwood tree-induced light

gradient. Agrofor Syst 75:189–196. doi:10.1007/s10457-

008-9175-9

Gaussoin RE, Baltensperger AA, Coffey BN (1988) Response

of 32 bermudagrass clones to reduced light intensity.

HortScience 23:178–179

Gold MA, Garrett HE (2009) Agroforestry nomenclature, con-

cepts, and practices. In: Garrett HE (ed) North American

agroforestry: an integrated science and practice, 2nd edn.

American Society of Agronomy Inc, Madison, pp 45–56

Jiang Y, Duncan RR, Carrow RN (2004) Assessment of low

light tolerance of seashore paspalum and bermudagrass.

Crop Sci 44:587. doi:10.2135/cropsci2004.5870

Kephart KD, Buxton DR, Taylor ES (1992) Growth of C3 and

C4 perennial grasses under reduced irradiance. Crop Sci

32:1033–1038. doi:10.2135/cropsci1992.0011183X00320

0040040x

Lin CH, McGraw RL, George MF, Garrett HE (1999) Shade

effects on forage crops with potential in temperate agro-

forestry practices. Agrofor Syst 44:109–119. doi:10.1023/

A:1006205116354

Loach K (1970) Shade tolerance in tree seedlings. II. Growth

analysis of plants raised under artificial shade. New Phytol

69:273–286. doi:10.1111/j.1469-8137.1970.tb02426.x

Martsolf JD (1966) Microclimatic modification through shade

induced changes in net radiation. Ph.D. Dissertation,

University of Missouri-Columbia

Pearson HA (1983) Forest grazing in the southern United States.

In: Hannaway DB (ed) Proceedings of the international

hilllands symposium foothills for food and forests. Cor-

vallis, OR, USA, pp 247–260

Peri PL, McNeil DL, Moot DJ, Varella AC, Lucas RJ (2002) Net

photosynthetic rate of cocksfoot leaves under continuous

and fluctuating shade conditions in the field. Grass Forage

Sci 57:157–170. doi:10.1046/j.1365-2494.2002.00312.x

Semchenko M, Lepik M, Gotzenberger L, Zobel K (2012)

Positive effect of shade on plant growth: amelioration of

stress or active regulation of growth rate? J Ecol

100:459–466. doi:10.1111/j.1365-2745.2011.01936.x

USDA NRCS (2016) The PLANTS Database. National Plant

Data Team, Greensboro, NC 27401-4901 USA. http://

plants.usda.gov. Accessed 25 April 2016

Valladares F, Niinemets U (2008) Shade tolerance, a key plant

feature of complex nature and consequences. Annu Rev

Ecol Evol Syst 39:237–257. doi:10.1146/annurev.ecolsys.

39.110707.173506

Valladares F, Sanchez-Gomez D, Zavala MA (2006) Quantita-

tive estimation of phenotypic plasticity: bridging the gap

between the evolutionary concept and its ecological

applications. J Ecol 94:1103–1116. doi:10.1111/j.1365-

2745.2006.01176.x

Van Sambeek JW, Navarrete-Tindall NE, Garrett HE, Lin CH,

McGraw RL, Wallace DC (2007) Ranking the shade tol-

erance of forty-five candidate groundcovers for agro-

forestry plantings. Temp Agroforester 15. http://www.nrs.

fs.fed.us/pubs/jrnl/2007/nrs_2007_vansambeek_002.pdf

Varella AC, Moot DJ, Pollock KM, Peri PL, Lucas RJ (2010) Do

light and alfalfa responses to cloth and slatted shade rep-

resent those measured under an agroforestry system?

Agrofor Syst 81:157–173. doi:10.1007/s10457-010-9319-6

Watson VH, Hagedorn C, Knight WE, Pearson HA (1984)

Shade tolerance of grass and legume germplasm for use in

the southern forest range. J Range Manag 37:229–232.

doi:10.2307/3899143

Wong CC, Rahim H, Sharudin MAM (1985a) Shade tolerance

potential of some tropical forages for integration with

plantations. 1. grasses. Mardi Res Bull 13:225–240

Wong CC, Sharudin MAM, Rahim H (1985b) Shade tolerance

potential of some tropical forages for integration with

plantations. 2. legumes. Mardi Res Bull 13:249–269

Yang W, Liu F, Zhou L, Zhang S, An S (2012) Growth and

photosynthetic responses of Canarium pimela and

Nephelium topengii seedlings to a light gradient. Agrofor

Syst 87:507–516. doi:10.1007/s10457-012-9570-0

Agroforest Syst

123